operation.h File Reference

Operation node can generate one or multiple Tensors. More...

#include <tvm/arith/analyzer.h>
#include <tvm/ffi/reflection/registry.h>
#include <tvm/ir/cow.h>
#include <tvm/ir/prim/expr.h>
#include <tvm/te/tensor.h>
#include <tvm/tirx/buffer.h>
#include <tvm/tirx/op.h>
#include <string>
#include <unordered_map>
#include <utility>
#include <vector>

Include dependency graph for operation.h:

This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Classes
class	tvm::te::CommReducerNode
	A commutative reducer node to represent a commutative binary operator with identity element. More...
class	tvm::te::CommReducer
	Managed reference to CommReducerNode. More...
class	tvm::te::ReduceNode
	Reduction operator. More...
class	tvm::te::Reduce
	Managed reference to ReduceNode. More...
struct	tvm::te::TensorDom
	Temporary data structure to store union of bounds of each axis of Tensor. More...
class	tvm::te::OperationNode
	Base class of all operation nodes. More...
class	tvm::te::PlaceholderOpNode
	A placeholder op represents an input placeholder. More...
class	tvm::te::PlaceholderOp
	Managed reference to PlaceholderOpNode. More...
class	tvm::te::BaseComputeOpNode
	A Compute op that compute a tensor on certain domain. This is the base class for ComputeOp (operating on a scalar at a time) More...
class	tvm::te::ComputeOpNode
	A Compute op that compute a tensor on certain domain. More...
class	tvm::te::ComputeOp
	Managed reference to ComputeOpNode. More...
class	tvm::te::ScanOpNode
	Symbolic scan. More...
class	tvm::te::ScanOp
	Managed reference to ScanOpNode. More...
class	tvm::te::ExternOpNode
	External computation that cannot be splitted. More...
class	tvm::te::ExternOp
	Managed reference to ExternOpNode. More...

Namespaces
namespace	tvm
	An object that builds and maintains block scope and StmtSref mapping for Dependence analysis.
namespace	tvm::te
	Tensor expression language DSL.
namespace	tvm::ffi

Typedefs
using	tvm::te::FCompute = std::function< PrimExpr(const ffi::Array< PrimVar > &i)>
	The compute function to specify the input source of a Tensor.
using	tvm::te::FBatchCompute = std::function< ffi::Array< PrimExpr >(const ffi::Array< PrimVar > &i)>
	The compute function to specify the inputs source of Tensors.

Functions
PrimVar	tvm::te::var (std::string name_hint, PrimType t=PrimType::Int(32))
	Construct a new Var expression.
IterVar	tvm::te::thread_axis (Range dom, std::string tag)
	Create a new IterVar that represents an axis in thread.
IterVar	tvm::te::reduce_axis (Range dom, std::string name="rv")
	Create a new IterVar for reduction operations.
Tensor	tvm::te::placeholder (ffi::Array< PrimExpr > shape, PrimType dtype=PrimType::Float(32), std::string name="placeholder")
	create a place holder tensor.
Tensor	tvm::te::compute (ffi::Array< PrimExpr > shape, FCompute fcompute, std::string name="tensor", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})
	Construct a new tensor by computing over shape, using the computation rule: result_tensor[axis] = fcompute(axis)
ffi::Array< Tensor >	tvm::te::compute (ffi::Array< PrimExpr > shape, FBatchCompute fcompute, std::string name="tensor", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})
	Construct a new tensor by computing over shape, using the computation rule: result_tensor[axis] = fcompute(axis)
ffi::Array< Tensor >	tvm::te::scan (ffi::Array< Tensor > init, ffi::Array< Tensor > update, ffi::Array< Tensor > state_placeholder, ffi::Array< Tensor > inputs=ffi::Array< Tensor >(), std::string name="scan", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})
	Construct new tensors by scan.
Tensor	tvm::te::compute (ffi::Array< PrimExpr > shape, std::function< PrimExpr(PrimVar)> f, std::string name="tensor", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})
Tensor	tvm::te::compute (ffi::Array< PrimExpr > shape, std::function< PrimExpr(PrimVar, PrimVar)> f, std::string name="tensor", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})
Tensor	tvm::te::compute (ffi::Array< PrimExpr > shape, std::function< PrimExpr(PrimVar, PrimVar, PrimVar)> f, std::string name="tensor", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})
Tensor	tvm::te::compute (ffi::Array< PrimExpr > shape, std::function< PrimExpr(PrimVar, PrimVar, PrimVar, PrimVar)> f, std::string name="tensor", std::string tag="", ffi::Map< ffi::String, ffi::Any > attrs={})

Detailed Description

Operation node can generate one or multiple Tensors.